stering corticosterone in the drinking water of female Swiss Webster mice for 17 or 18 days (13
mg/kg). Light dark emergence, startle habituation, and startle reactivity were measured. Chronic but not
acute treatment with corticosterone increased the latency to emerge into the light compartment, an
anxiogenic-like effect. Chronic corticosterone treatment did not affect startle habituation, but did reduce
startle reactivity. This study suggests that chronic hypercortisolemia may contribute to anxiety-related
behavior in patients with Cushings disease and depression.
Keywords: depression, cortisol, anxiety
Prolonged hypersecretion of hypothalamicpituitaryadrenal
(HPA) axis effectors, such as corticotropin-releasing hormone
(CRH) and cortisol, in individuals with depression is one of the
most highly replicated findings in biological psychiatry (Nemeroff
& Evans, 1984; Sachar et al., 1985; Varghese & Brown, 2001).
However, despite intensive investigation, it is not yet clear whether
there is a causal relationship between HPA axis dysfunction and
psychiatric liability (Gillespie & Nemeroff, 2005; Sapolsky, 2000).
Understanding whether HPA axis effectors can induce some of the
symptoms of depression is critical not only for understanding the
etiology of the disease, but also for identifying novel therapeutic
targets for treating depression. Therefore, determining the precise
relationship between elevated HPA axis effectors and psychiatric
illness is of great importance.
At least four lines of evidence suggest that corticosteroids may
actually cause the mood and behavior changes in depression and,
specifically, the anxiety-like features that are frequently comorbid
with depression. First, more than 50% of individuals with Cush-
ings disease, characterized by high blood cortisol levels, present
with symptoms of depression and anxiety (Fava, Sonino, & Mor-
phy, 1987; Kelly, 1996; Sablowski, Pawlik, Ludecke, & Herr-
mann, 1986). Second, the anxious-retarded subtype of depression,
characterized by high anxiety levels and psychomotor retardation,
is commonly associated with disruption of the HPA axis (de
Winter et al., 2003). Third, individuals receiving glucocorticoid
therapy for inflammatory and other disorders have long been
known to have an increase in mood-related side effects, including
anxiety and depression (Kayani & Shannon, 2002; Soliday, Grey,
& Lande, 1999). Finally, elevated glucocorticoid levels for chronic
periods are associated with increased activity in anxiety-related
brain regions, such as the amygdala, in both rodents and humans
(Drevets et al., 2002; Erickson, Drevets, & Schulkin, 2003;
Makino, Gold, & Schulkin, 1994a, 1994b; Schulkin, Gold, &
McEwen, 1998). On the basis of this evidence, we hypothesized
that prolonged but not acute glucocorticoid treatment would in-
crease anxiety-like behavior in the light dark emergence task.
Because the startle response is another unconditioned behavior that
can be modulated by affective states (Grillon & Baas, 2003), we
also tested the effects of acute and chronic corticosterone on
habituation of the startle response and overall startle reactivity.
The idea that chronic but not acute glucocorticoid treatment may
selectively affect anxiety-like behavior fits with the large literature
suggesting that acute and transient stress system activation is
adaptive and helps to maintain homeostasis (Munck, Guyre, &
Holbrook, 1984). This is in contrast to the effects of chronic stress
system activation or chronic exposure to HPA axis effectors,
which may be maladaptive or increase vulnerability to psychiatric
disease (McEwen, 2003; Schulkin, McEwen, & Gold, 1994). We
report here that chronic but not acute administration of corticoste-
rone to mice by using a noninvasive treatment method resulted in
an anxiogenic-like effect in the mouse light dark box (Crawley,
1981; Malmberg-Aiello, Ipponi, Bartolini, & Schunack, 2002).
Chronic corticosterone, however, did not affect startle habituation
but did reduce startle reactivity.
Materials and Method
Subjects
Adult (30 35 g) female Swiss Webster mice (Taconic Farms, German
Town, NY) were group housed (5 per cage) in plastic cages in a
temperature- and light-controlled room. Mice had access to food and water
ad libitum. Mice were habituated to the facility for at least 1 week before
Paul Ardayfio and Kwang-Soo Kim, Laboratory of Molecular Neuro-
biology, Mclean Hospital, and Program in Neuroscience, Harvard Medical
School.
This work was supported by grants from the National Institutes of
Health (MH 48866 and 400042, DC 006501) and by awards from the
National Alliance for Research on Schizophrenia and Depression.
Correspondence concerning this article should be addressed to Kwang-
Soo Kim, Mclean Hospital, 115 Mill Street, MRC 215, Belmont, MA
02478. E-mail: kskim@mclean.harvard.edu
Behavioral Neuroscience
Copyright 2006 by the American Psychological Association
2006, Vol. 120, No. 2, 249 256
0735-7044/06/$12.00
DOI: 10.1037/0735-7044.120.2.249
249 testing. All procedures were conducted in accordance with Mclean Hos-
pital institutional animal care and use committee-approved protocols.
Drugs
A well-established, noninvasive corticosterone treatment method (Fair-
child, Leitch, & Ingram, 2003; Magarinos, Orchinik, & McEwen, 1998;
Nacher, Gomez-Climent, & McEwen, 2004) was used so as to avoid the
stress associated with chronic intraperitoneal injections. Corticosterone
(Sigma, St. Louis, MO) was dissolved in ethanol, diluted, and administered
via the drinking water at a final concentration of 0 or 35
g/ l in 0.3%
ethanol. This concentration resulted in the ingestion of approximately 13
mg/kg per day of corticosterone. Mice in the chronic treatment group were
treated for 17 or 18 days with corticosterone or vehicle containing 0.3%
ethanol in the drinking water. On Day 17, mice were tested for anxiety-like
behavior in the light dark emergence task. On Day 18, a different set of
mice, also receiving corticosterone, were tested for startle behavior. Mice
in the acute treatment group received a 24-hr exposure to corticosterone in
the drinking water. Pentobarbital (15 mg/kg) was dissolved in 0.9% saline
and was injected intraperitoneally 15 min before light dark testing.
LightDark Emergence
The light dark box test makes use of rodents natural aversion to bright
areas compared with darker ones. In the two-compartment light dark box,
rodents prefer the smaller dark area and hesitate to enter the brightly lit,
open area. Although there have been numerous adaptations of the light
dark box (Crawley, 1985; Imaizumi, Miyazaki, & Onodera, 1994; Onaivi
& Martin, 1989; Stratton et al., 1993), the basic premise always remains the
conflict between exploration of a large, bright, and open area and the safety
of the smaller, enclosed area. In our preliminary studies, we found that
when started in the light side of the compartment, mice given anxiogenic
treatments did not explore rapidly enough to find and enter the dark
compartment; instead, they tended to freeze and remain immobile for a
majority of the test session (Ardayfio & Kim, 2004). This resulted in a
number of false positives because increased latencies to enter the dark
compartment are generally thought to reflect anxiolytic-like effects. In-
stead, we found the emergence latency to leave the dark compartment and
enter the light compartment to be the most reliable indicator of anxiety-like
behavior sensitive to both anxiogenic and anxiolytic treatments. The emer-
gence task was carried out in a plastic apparatus (56 cm long
33 cm wide
30 cm deep), which was divided into two compartments by a vertical
sliding door that remained open (8 cm). The larger exploration compart-
ment comprising two thirds of the apparatus was transparent, open, and
illuminated by a 60-W lamp placed above the compartment. The smaller
start compartment was black and had a lid that was closed during testing.
At the beginning of each session, the mouse was placed in a far corner of
the dark box, facing the light compartment. The latency to enter the light
compartment with all four paws was recorded. Mice that failed to enter the
lit compartment within 10 min were removed and given a maximum
latency score of 10 min. The floor of each box was cleaned with 70%
ethanol between sessions, and mice were tested in a counterbalanced order
in regard to treatment.
Home-Cage Return Latency
We developed a novel home-cage return task as a control measure to
ensure that long latencies to emerge from the dark box did not reflect a
general inability to locomote, but instead reflected anxiety-related behav-
ior. The apparatus used was a Plexiglas runway (47 cm long
5 cm wide
21 cm deep) that was open at one end. The home cage of test animals
was turned on the side and placed at the open end of the runway such that
mice could ambulate from the runway directly into the cage. Mice were
individually placed at the closed end of the runway and allowed to traverse
the runway into the home cage. The time required to traverse the runway
and place all four paws in the home cage was recorded as the latency.
Startle Habituation
Startle reflexes were measured by using the San Diego Instruments (San
Diego, CA) SR-Lab system, consisting of a nonrestrictive Plexiglas cyl-
inder (4-cm inner diameter, 13-cm length) mounted on a Plexiglas platform
and placed in a ventilated, sound-attenuated chamber. Cylinder movements
were detected and measured by a piezoelectric element mounted under
each cylinder. Chambers were calibrated before use